Stromal-epithelial interaction was demonstrated to affect androgen responsiveness and the growth and progression of prostate cancer. In a cellular model of prostate carcinogenesis, we demonstrated that organ- specific stroma, bone and prostate fibroblasts induced LNCaP tumor growth in athymic mice and accelerated its progression to androgen-independent and metastatic state. Specific objectives of this proposal are: Specific Aim 1 proposes to characterize cytogenetic profiles of parental human LNCaP and its androgen-independent and metastatic sublines derived from chimeric tumors under the influence of fibroblasts derived from either transition or peripheral zone. The hypothesis to be tested is that zonal differences in fibroblasts determine the cytogenetic profile of the tumor and predict its malignant potential. Specific Aim 2 seeks to define the roles of paracrine/autocrine mediators, transforming growth factor beta (TGFbeta)/TGFbeta receptors and hepatocyte growth factor/scatter factor (HGF/SF)/c-met protooncogene pathways in clonal interaction between androgen-dependent (AD) and androgen-independent (AI) prostate cancer cells. DNA-mediated gene transfer of sense, antisense, or dominant negative constructs of growth factor genes and delivery of selective growth factor and receptor agonists and antagonists will be employed to evaluate clonal interactions between AD and AI cells. The hypothesis to be tested is that by interrupting specific growth factor and receptor interactions, the clonal balance can be tipped such that the rate of tumor progression could be affected. Specific Aim 3 explores the action of two paracrine mediators, keratinocyte growth factor (KGF) and HGF/SF, in models of normal and regenerative growth of the rat prostate gland. We will study the effect of continuous systemic infusion of KGF or HGF/SF on the growth of accessory sex organs and determine the action of androgen on KGF or HGF/SF mRNA and protein expression. Paracrine action of these growth factors in promoting differentiation, apoptosis, and/or senescence of prostate epithelium will be emphasized. The implications of this study are that cytogenetic profiles of cancer epithelium may be dictated in part by host stroma. The study of genetic alterations in cells with tight cell- lineage relationship is most likely to reveal markers indicative of disease progression. Paracrine/autocrine mediator pathways are involved in clonal interactions between AD and AI cells in prostate cancer, and can be interrupted to impede cancer progression. In addition to growth, KGF and HGF/SF may have profound action in stimulating prostate cancer differentiation, apoptosis, and/or senescence.